1 Introduction

Berberine is an isoquinoline alkaloid present in many Berberidaceae plants, and its hydrochloride possess broad spectrum of anti-microbial effect for the treatment of intestinal infection and diarrhea in extensive clinical application (Taylor and Baird 1995). It is considered to be the major bioactive component of various traditional Chinese medicines, such as Coptis Chinensis Franch and Cortex Phellodendri Chinensis (Zhang et al. 2010). In recent years, many studies have also confirmed its anti-diabetes, anti-tumor and anti-arrhythmia effect (Sanchez-Chapula 1996; Iizuka et al. 2000; Chen and Chang 1995), and clinical trials are being conducted to evaluate its potential ability to different human diseases.

The metabolism and pharmacokinetics of berberine have been studied using HPLC with ultraviolet or mass spectrometry detectors, and the former proved to be a very efficient method. Berberine was detected at a very low concentration in serum after oral administration, and P-glycoprotein (P-gp) was considered as the major factor to its low bioavailability in vivo. In-vitro Caco-2 cell model indicated that the intestinal absorption of berberine might reduced by P-gp through drug efflux function (Feng et al. 2006). However, in vitro models could not comprehensively exhibit the absorption ability of berberine, the study of blood concentration could provide more reliable details for pharmacokinetic characteristics of drug, especially the in vivo study of portal-vein model, which could elucidate the absorption kinetics of drug through intestinal absorption path.

Moreover, it is reported that P-gp inhibitor such as verapamil could enhance the intestinal absorption of drug by depressing the P-gp level (Wu et al. 2011). Borneol, which belongs to a traditional Chinese medicine, was also considered as a potential P-gp inhibitor, and intestinal perfusion study (in situ) indicated that borneol could improve the intestinal absorption of berberine in rat (Chen and Wang. 2003).

The effect of P-gp inhibitor was highly critical to absorption of low bioavailability drug such as berberine, and the improvement of verapamil and borneol to its absorption remains unclear on pharmacokinetic level. This paper aims to investigate the portal-vein absorption kinetics of berberine and the effect of verapamil and borneol to its absorption ability in vivo. Moreover, a fully validated UHPLC method for detecting berberine in blood sample and a portal-vein absorption kinetics model for pharmacokinetic studies were also established.

2 Materials and methods

2.1 Chemicals and reagents

Berberine, borneol, verapamil and nitidine chloride were purchased from National Institutes for Food and Drug Control. Sodium dodecyl sulfate, potassium dihydrogen phosphate and sodium carboxymethyl cellulose (CMC-Na) were purchased from Beijing Chemical Corporation (Beijing, China). HPLC grade acetonitrile was used for UHPLC analysis. De-ionized water was purified by a Milli-Q system (Millipore, Bedford, MA, USA). All the reagents were of analytical grade. The chemical structures of berberine, verapamil and nitidine chloride were shown in Fig. 1.

Fig. 1
figure 1

Chemical structures of berberine, verapamil and nitidine chloride (I.S)

Full size image

2.2 Animal

Male and female Sprague–Dawley rats of SPF grade were obtained from Laboratory Animal Center of Sichuan Health Science Academy (Certificate No. SCXK 2004-6). The rats were housed in an animal room in a controlled condition (22–24 °C) with free access to food and de-ionized water. All animal treatments were in accordance with the National Institutes of Health Guide to the Care and Use of Laboratory Animals. The experiments were carried out under the approval of Experiment Administration Committee of Sichuan Health Science Academy.

2.3 Drug administration

Berberine, verapamil and borneol were dissolved in 0.5 % CMC-Na solution, respectively, and the final concentration was 2.8 mg/mL for Berberine hydrochloride, 5.0 mg/mL for verapamil hydrochloride and 1.5 mg/mL for borneol.

The dosage of berberine and borneol was determined as 94.5 and 31.5 mg/kg according to Chinese Pharmacopoeia (2010 Edition) by surface area method of human dosage. Previous study had reported that intragastric administration of verapamil (20–100 mg) could significantly inhibit the Pg protein expression in rat intestinal tract (Cao et al. 2007). In the current research, we adjust the dosage to 50 mg/kg to prevent serious respiratory depression of animals.

2.4 Animal experiment and drug administration

Twenty four male and female rats weighed 180–200 g were divided into four groups (three male and three female rats of each group) randomly as follow: group berberine group (BG), berberine was administered individually by injecting berberine suspension into intestinal cavity; group verapamil + berberine group (VBG), berberine was administered according to the method of group BG after oral administration of verapamil; group borneol + berberine group (BBG), berberine was administered according to the method of group BG after oral administration of borneol. Moreover, group long-term use of borneol + berberine group (LBBG) was also established to evaluate the absorption kinetics of berberine after long-term administration of borneol.

For each group, 0.3 mL blood (n = 6) was collected before and after administration at 0, 15, 30, 60, 90, 120, 180 and 240 min. The blood samples were placed in heparinized tubes and immediately centrifuged to obtain the plasma. The rats in BG group were injected berberine suspension into intestinal cavity, containing drug 94.5 mg kg−1, from duodenum near pylorus when it has been stable for 10–15 min after anaesthesia and the surgery of portal-vein separation and intubation.

In VBG group, the rats received 0.1 % CMC-Na by gavage, suspension containing 50 mg kg−1 of verapamil. Then, the rats were performed the surgery of portal-vein separation and intubation under anesthesia 2 h later. At last, berberine suspension was given to each rat in the same way as that of BG group. The rats in BBG group received 0.1 % CMC-Na by gavage, suspension containing 31.5 mg kg−1 borneol. The next steps are the same as those of VBG group. The rats in LBBG group had been received 0.1 % CMC-Na by gavage, suspension containing 31.5 mg kg−1 borneol for 7 days. In the 8th day, all steps are exactly the same as those of BBG group.

2.5 Stock solutions, working solutions, and quality control (QC) samples

The standard stock solutions of berberine and nitidine chloride (IS) stock solutions were individually prepared in methanol to a final concentration of 163.20 and 50.60 μg/mL, respectively. Working solutions of berberine were prepared by further diluting the stock solutions with methanol to 326.4, 261.12, 163.2, 81.6, 32.64, 16.32, 10.20, 8.16 ng/mL. The working solution of IS was prepared by diluting the stock solution of nitidine chloride with methanol to 4 μg/mL. All the working solutions were stored at −20 °C.

Calibration curves were prepared by spiking 100 μL of blank plasma with 50 μL of the above mentioned berberine working solutions and 10 μL of the IS working solution. Samples were then vortex mixed for 60 s and extracted with 1,000 μL acetonitrile by vortex mixing for 1 min and ultrasonication for 30 s. After centrifugation at 12,000 rpm for 5 min, the upper organic layer was transferred to another tube and evaporated to dryness under a gentle stream of nitrogen at 20 °C. The residue was reconstituted in 100 μL mobile phase followed by vortex-mixing for 1 min and ultrasonication for 30 s. After centrifugation at 13,500 rpm for 10 min, a 50 μL aliquot of supernatants was injected into the UHPLC system for analysis. The linear concentration ranges were 4.08–163.20 ng/mL for berberine.

The quality control (QC) samples were prepared by adding the stock solution of berberine into blank plasma to obtain final concentrations of 5.10, 40.80 and 130.56 ng/mL, which represented low, medium, and high concentration QC samples, respectively.

2.6 UHPLC chromatographic conditions

Quantitative analysis was carried out on an Agilent UHPLC system. The UHPLC system consisted of a Agilent 1290 Infinity LC instrument (Agilent, CA, USA) equipped with an a quaternary solvent delivery system and an autosampler and a MWD detector. An Ultimate X13-C18 column (100 mm × 2.1 mm i.d., 3 μm) equipped with an Agilent Zorbax SB-C18 guard column (12.5 mm × 4.6 mm i.d., 5 μm) was adopted as solid phase. The mobile phase was optimized and consisted of acetonitrile–deionized water–potassium dihydrogen phosphate–sodium dodecyl sulfate (500 mL:500 mL:3.4 g:1.7 g) with a flow rate of 0.4 mL/min. The MWD detector was operated at 346 nm and the column temperature was 40 °C.

2.7 Sample preparation

To a 100 μL aliquot of plasma sample, 10 μL of the IS working solution and 40 μL KH2PO4 (1 M) were added. The other treatments were the same as described before in Sect. 2.5.

2.8 Data analysis

All the calculations were processed with MAPLE 15 software to get the parameters using statistical moment method. The major dynamic parameters of every group were tested and compared by t test with those of BG group performing with Statistical Tools of MAPLE 15.

3 Results and discussion

3.1 Method validation

3.1.1 Assay specificity

Nitidine chloride was chemically stable, and has similar chemical structure and chromatographic behavior with berberine. In this study, nitidine chloride was chosen as the internal standard for quantitation of berberine. The specificity was evaluated by comparing blank, spiked and routinely prepared rat plasma samples. The chromatogram of blank plasma did not interfere as observed in matrix at the retention times of berberine (3.21 min) and nitidine chloride (3.42 min), respectively. The representative UHPLC chromatograms of blank plasma sample were shown in Fig. 2a; blank plasma sample spiked with nitidine chloride and berberine hydrochloride was shown in Fig. 2b; plasma sample of 30 min after administration of berberine from BG group was shown in Fig. 2c.

Fig. 2
figure 2

Representative UHPLC chromatograms of berberine (Ber) and nitidine chloride (NC): a blank plasma; b blank plasma spiked with analytes and IS; c plasma sample obtained 15 min after administration of berberine from BG group

Full size image

3.1.2 Calibration curves and limit of detection

Calibration curves were established by analyzing a series of spiked samples as described under Sect. 2.5. The calibration curves (y = ax + b) were constructed by plotting peak area ratios of berberine to IS (y) against analyte concentrations (x), and 1/x 2 weighting power was applied. The regression equation for berberine was calculated as y = 0.1158x − 0.1509 (r = 0.9988) with a wide concentration ranges from 4.08 to 163.20 ng/mL. The limit of detection (LOD) was determined at a signal-to-noise ratio (S/N) of 3:1. The lower limit of quantitation (LLOQ) was determined as the lower end of the calibration range. The LOD and LLOQ of berberine were determined as 0.8 and 4.08 ng/mL, respectively.

3.1.3 Precision and accuracy

Precision was determined by analyzing three different concentrations of QC samples (six replicates for each concentration) for three consecutive days. Within- and between-batch precisions were evaluated by relative standard deviation (RSD), which ranged from 2.21 to 7.58 % (data shown in Table 1). Accuracy was calculated as relative error (RE) by the following formula: RE % = [(C det − C nom)/(C nom)] × 100, where C nom represented the nominal concentration and C det represented the mean value of detected concentrations. The calculated accuracy values for berberine were −15.8 to 3.59 % (data shown in Table 1). The above data met the requirements for bioanalytical method validation, indicating that this method was acceptable for pharmacokinetic studies.

Table 1 Precision, accuracy and recovery of berberine
Full size table

3.1.4 Recovery

Extraction recoveries were determined by comparing routinely prepared QC samples with routinely prepared blank samples spiked with QC solutions. Recoveries were calculated by the formula: recovery (%) = concentration found/concentration spiked × 100 %. Extraction recoveries determined for berberine after protein precipitation ranged from 85.34 to 111.62 % (Table 1). The extraction recovery of IS was above 85 %. The extraction recoveries were acceptable for bioanalysis.

3.1.5 Stability

Stability of berberine in plasma was tested at three QC concentration levels and by a series of experiments, (a) freeze and thaw stability, stability of plasma after three cycles of freezing at −80 °C and thawing at room temperature; (b) short-term room temperature stability, stability of plasma at room temperature for 2 h during sample pretreatment process; (c) long-term stability, stability of plasma at −80 °C for 1 month; and (d) post-preparation stability, stability of the extracted samples in autosampler (10 °C) for 24 h. In addition, the stability of berberine and IS in stock and working solutions at 1–4 °C for 1 month was also evaluated. The berberine and IS were proved stable both in methanol (at −4 °C) and in plasma (at −80 °C) after long-term storage for 1 month. Degradation of the analytes after three freeze–thaw cycles, at room temperature storage (2 h), or at autosampler storage (24 h) was below 15 and 10 %, respectively. These data indicated that all the analytes were stable during sample preparation and chemical analyses.

3.1.6 Pharmacokinetic and PK parameters analysis

The established UHPLC method was applied to the quantitative determination of berberine concentrations in plasma of differnet groups after portal-vein administration. The calculated concentrations of each time are shown in Table 2. The pharmacokinetic parameters are shown in Table 3 and the representative elimination curves of berberine are also depicted in Fig. 3.

Table 2 Plasma concentrations of berberine in different groups after portal-vein absorption (n = 6)
Full size table
Table 3 Portal-vein absorption kinetic parameters of berberine in plasma of different groups
Full size table
Fig. 3
figure 3

Plasma concentration–time profiles of berberine in portal-vein absorption rat model of group BG, group VBG, group BBG and group LBBG, respectively

Full size image

By examining 15–240 min plasma concentration of berberine, we found that there was no remarkable change of AUC value by comparing group VBG, BBG and LBBG to group BG (P > 0.05, data were shown in Table 3), which indicating total absorption of berberine was not affected by either co-administration of verapamil or borneol. However, significant changes were observed from T max among group BBG, group LBBG and group BG (P < 0.05, see Table 3), which indicating co-administration of berberine with borneol in advance could shorten the T max. Interestingly, the opposite phenomenon was observed in group VBG. After co-administration of verapamil, the T max was prolonged comparing to single dosage of berberine (group BG), but the differences between two groups were not significant. Meanwhile, we found that the C max value of group BBG and group LBBG were also remarkably reduced comparing to group BG (P < 0.05, see Table 3). Based on the results above, it seems that co-administration of borneol in advance through single using or long term using could defer the whole absorption process of berberine, as in both group BBG and group LBBG, T max and C max of berberine were significantly prolonged and reduced, respectively, while the total absorption was stable. However, co-administration of verapamil seems had little influence to berberine’s absorption ability.

Previous study proved that the pharmacokinetic absorption of berberine exhibited notable individual differences in both animal and human experiments after gastrointestinal administration (Hu et al. 2009). Similar result was also observed in our current study, we found that AUCmax value of berberine was detected as 5 times higher than AUCmin value, and the ratio of standard error to the mean (mean of AUC value) reached almost 54 % within group BG. These evidences together supported the previous report, and indicated significant individual absorption differences after single administration of berberine. Interestingly, after co-administration of verapamil, borneol and long-term co-administration of borneol for 7 days in a row, the ratio of standard error and the mean reduced to 50 % within group VBG, 24 % within group BBG and 18 % within group LBBG, respectively, which meant the individual absorption differences reduced. Moreover, C max of other three groups was also remarkably reduced as compared to group BG (P < 0.05). Taken together, it appears that co-administration of borneol could eliminate the individual absorption differences and smooth the absorption curves without affecting the total absorption, especially through long-term administration. Based on this preliminary research, the mechanism for co-existing borneol to change the pharmacokinetic process warrants further studying next.

As typical Pg inhibitors, verapamil did not exhibit the ability to improve the absorption of berberine but to reduce its total absorption, which is opposite to the result of cellular transport study (Cao et al. 2007). Besides the fluctuation of AUC values caused by individual absorption differences, it is possible that the expression of Pg protein in gastrointestinal tract and the affinity of berberine to Pg protein will also lead to this result. It is reported that Pg protein in rat is translated and expressed by Abcb1b, and highly enriched in placenta, liver and brain tissues while lowly expressed in gastrointestinal tract (Yang et al. 2010). Moreover, we speculated that the affinity of berberine to Pg protein was not as high as we expected, for it had not yet been studied, so far. Taken together, this may explain the reason why the total absorption of berberine is reduced in our study based on portal-vein absorption model as compared to the previous study based on Caco-2 cellular model, in which Pg protein is highly expressed.

It is reported that co-existing borneol in animal experiment could change the pharmacokinetic absorption, and administration of borneol in advance could reduce the T max significantly (Yuan and Li 2010). In our study, similar result was observed. Berberine was administrated to rats after using borneol for 2 h in group BBG and group LBBG, and we found that T max of both group was reduced. However, interestingly, previous report also indicates that co-administration of drugs with borneol synchronously usually delay the T max (Chen et al. 1990; Xiao et al. 2009). It is speculated that when testing-drugs and borneol were administrated together synchronously, the borneol must be absorbed first and then accelerate the absorption of testing-drugs, which may cause the delay in the T max. Based on our results and the former studies, we found that the absorption-accelerated ability of borneol normally depended on its administrated time. Herein, we suggest that borneol should be administrated in advance before therapy drugs in clinical application to exhibit its absorption-accelerated ability.

4 Conclusion

In the present study, a portal-vein absorption model was established for pharmacokinetic study of berberine in rats by a fully validated UHPLC method. The absorption and interaction of berberine and verapamil, borneol were studied for the first time using the established model. We found that single dosage of verapamil in advance had no remarkable improvement to berberine’s portal-vein absorption. Moreover, it is clearly demonstrated that co-administration of borneol 2 h in advance could significantly reduce the T max and C max, which also delay the absorption process. This result illustrates the pharmacokinetic behavior of berberine in plasma based on portal-vein model, and maybe helpful to explain the synergies mechanism of co-administration of berberine with P-gp inhibitors.